The paper presents a methodology for the structural and parametric synthesis of neural network controllers used in automatic control systems. The approach uses artificial neural networks to create controllers that are integrated into a control system with control objects. The synthesis of the neural network controller is achieved using reinforcement learning and pre-building a neural network imitator of the control object. This method is effective when classical control system synthesis methods are not applicable due to significant nonlinearity and the difficulty in forming a mathematical model of the control object with the required accuracy.
Publication date: 29 Dec 2023
Project Page: Not Provided
Paper: https://arxiv.org/pdf/2312.16510
